llvm-svn: 118650

invoke the test driver to rerun the very same test.  Example output:

/Volumes/data/lldb/svn/trunk/test $ tail 2010-11-09-14_51_34/ExpectedFailure-TestSettings.SettingsCommandTestCase.test_set_output_path.log 

Traceback (most recent call last):
  File "/Volumes/data/lldb/svn/trunk/test/settings/TestSettings.py", line 136, in test_set_output_path
    "'stdout.txt' exists due to target.process.output-path.")
AssertionError: False is not True : 'stdout.txt' exists due to target.process.output-path.


To rerun this test, issue the following command from the 'test' directory:

./dotest.py -v -t -f SettingsCommandTestCase.test_set_output_path

llvm-svn: 118646

in the type copy routine to make type problems
easier to debug.

llvm-svn: 118638

    @unittest2.skip("rdar://problem/8641483 ./dotest.py -v -t -w forward seg faults")

and add a @skip decorator for test_with_dwarf_and_run_command() method:

    @unittest2.skip("rdar://problem/8648070 'expression *bar_ptr' seg faults")

llvm-svn: 118632

be different than the class name which uses frame variable command.

llvm-svn: 118602

takes the form:

    "Command '%s' returns successfully" % str

and the latter takes the form:

    "'%s' returns expected result" % str
    or
    "'%s' matches expected result" % str

llvm-svn: 118599

llvm-svn: 118591

llvm-svn: 118590

llvm-svn: 118511

llvm-svn: 118489

to not get resolved.

Fixed the "void **isa_ptr" variable inside the objective C verifier to start
with a '$' character so we don't go looking for it in our program.

Moved the lookup for "$__lldb_class" into the part that knows we are looking
for internal types that start with a '$'.

llvm-svn: 118488

I only did a tiny bit of testing; in the one case I tried changing the
contents of a radar in the middle of a stack and it was still current in
the live register context so it filtered down to frame 0 and was handed
over to the live register set RegisterContext.  I need to test a case
where a register is saved on the stack in memory before I check this
one off.

llvm-svn: 118486

the frame count is requested or each frame is individually requested.

In practice this doesn't seem to help anything because we have
functions like StackFrameList::GetNumFrames() which is going to
request each frame anyway.  And classes like ThreadPlanStepRange
and ThreadPlanStepOverRange get the stack depth in their ctor forcing
a full stack walk.  But at least UnwindLLDB will delay doing a full
walk if it can.

llvm-svn: 118477

FuncUnwinders object if the eh_frame section was missing
from an objfile.  Worked fine on x86_64 but on i386 where
eh_frame is unusual, that resulted in the arch default 
UnwindPlan being used all the time instead of picking up
an assembly profile based unwindplan.

llvm-svn: 118467

'run'          => Control-C r (gud-run)            "Run the program."
'process kill' => Control-C s (gud-stop-subjoball) "Stop the program."

llvm-svn: 118460

llvm-svn: 118450

llvm-svn: 118438

failed/errored tests at the start of the test run, in case the test run crashes
for any reason.  That way, it is easy to locate the session logs for accumulated
failures/errors.

llvm-svn: 118427

llvm-svn: 118415

a pseudo terminal even when the process being attached to. 

Fixed a possible crasher in the in:

    bool
    ClangASTContext::IsAggregateType (clang_type_t clang_type);
    
It seems that if you pass in a record decl, enum decl, or objc class decl
and ask it if it is an aggregate type, clang will crash. 

llvm-svn: 118404

when a function starts and ends, and also the 
disassembly for anything that is a client of
ClangExpressionParser after it has been JIT
compiled.

llvm-svn: 118401

the same architecture as a default program, the attach architecture auto
detection would change the architecture to the architecture of the darwin-debug
(which was always x86_64) and hose up your debug session.

llvm-svn: 118399

declarations when they should have been.

llvm-svn: 118393

we want to run just the instance of testclass-testmethod combination and nothing else.

Specifying '-g' now will admit the whole .py test module if it does not contain a matching
testclass-testmethod combination at all.

This option arrangement adheres to the RISC principle of making the common cases fast. :-)

rdar://problem/8584914 Can the default for dotest.py's "-g" flag be switched?

llvm-svn: 118392

every external variable reference in the module,
and returning a clean error (instead of letting
LLVM issue a fatal error) if the variable could
not be resolved.

llvm-svn: 118388

and "/private/tmp/a.c". This was done by adding a "mutable bool m_is_resolved;"
member to FileSpec and then modifying the equal operator to check if the
filenames are equal, and if they are, then check the directories. If they are
not equal, then both paths are checked to see if they have been resolved. If
they have been resolved, we resolve the paths in temporary FileSpec objects
and set each of the m_is_resolved bools to try (for lhs and rhs) if the paths
match what is contained in the path. This allows us to do more intelligent
compares without having to resolve all paths found in the debug info (which
can quickly get costly if the files are on remote NFS mounts).

llvm-svn: 118387

where the DWARF is in the .o files so they can track down the actual type for
a forward declaration. This was working before for just DWARF files, but not
for DWARF in .o files where the actual definition was in another .o file.

Modified the main thread name in the driver to be more consistent with the
other LLDB thread names.

llvm-svn: 118383

don't crash if we disable logging when some code already has a copy of the
logger. Prior to this fix, logs were handed out as pointers and if they were
held onto while a log got disabled, then it could cause a crash. Now all logs
are handed out as shared pointers so this problem shouldn't happen anymore.
We are also using our new shared pointers that put the shared pointer count
and the object into the same allocation for a tad better performance.

llvm-svn: 118319

Howard Hinnant gave us changes for lldb_private::SharingPtr that gives us the ability have a single allocation contain both the class and the ref count without having to do intrusive pointer type stuff. They will intermingle correctly with other shared pointers as well. In order to take advantage of this you need to create your pointer in your class with the make_shared function:

       lldb_private::SharingPtr<A> p = llvm::make_shared<A>(i, j);

Currently up to five constructor arguments are supported and each must be an LValue.

llvm-svn: 118317

of store statements were not being marked for
resolution.

llvm-svn: 118316

identification of the test failures/errors by human beings as well as automatic
processings.

The prefix which identifies the type can be: Error, Failure, or ExpectedFailure.

llvm-svn: 118315

so we don't end up with weak exports with some compilers.

llvm-svn: 118312

pseudoterminal to pass to the inferior for the inferior's I/O
(to allow direct writing, rather than passing all the I/O around
via packets).

llvm-svn: 118308

    rdar://problem/8557095 lldb disas crashed (from lldb developer)

llvm-svn: 118299

Added the equivalent of gdb's "unwind-on-signal" to the expression command, and a parameter to control it in ClangUserExpression, and on down to ClangFunction.  

llvm-svn: 118290

llvm-svn: 118285

llvm-svn: 118284

llvm-svn: 118283

llvm-svn: 118282

llvm-svn: 118279